A significant need exists in advanced avionics applications for intelligent systems to aid pilots in making in-flight tactical decisions. Pilots are overwhelmed by the quantity of incomong data and communications upon which life-critical decisions are based. Confronted with numerous dynamically changing situations, pilots must precisely activate the dozens of switches, buttons, and knobs covering their controls. While each of the aircraft's hundreds of components and systems serve legitimate purposes, the technologies which created them have far outpaced the technology for intelligently interfacing these systems to the pilot. Traditional real-time expert system control strategies may fail to provide reliable responses in time-constrained avionics applications. This paper describes an inference strategy designed to operate in time-constrained military domains. Based upon the problem-solving paradigms which enable military decision making, this inference strategy considers the most influential knowledge first, and less significant data as time permitsThis time-constrained inference strategy utilizes a semantic network to represent domain knowledge and a time-merit value to guide the search. The certainty of the expert system response increases in proportion to the amount of processing time available. The applicability of this inference strategy for real-time avionics applications has been successfully demonstrated in a prototype expert system for weapon-to-threat assignment.
A time-constrained inference strategy for real time expert systems
Eine zeitbeschraenkte Entscheidungsstrategie fuer Echtzeit-Expertensysteme
1985
6 Seiten, 1 Bild, 7 Quellen
Conference paper
English
A Parallel Strategy for Implementing Real-Time Expert Systems Using CLIPS
British Library Conference Proceedings | 1995
|Real Time Analysis on Expert Systems
British Library Conference Proceedings | 2000
|Expert systems for real time applications
Tema Archive | 1985
|Real-time space system control with expert systems
NTRS | 1988
|